Introduction

The aging population has changed the daily work of hospital medicine. Older adults account for a large share of medical admissions, and many arrive with cognitive impairment, frailty, sensory impairment, polypharmacy, acute infection, dehydration, pain, immobility, sleep disruption, metabolic stress, or organ dysfunction. Each of these factors increases vulnerability to delirium.

Delirium is an acute disturbance in attention and awareness that develops over a short period of time and tends to fluctuate. It may be hyperactive, hypoactive, or mixed. Hypoactive delirium is especially easy to miss because patients may appear quiet, tired, withdrawn, or less engaged rather than agitated. Missing delirium matters because it is associated with poor outcomes and often signals an untreated precipitant.

Hospitalists are in a difficult position. The interventions most likely to prevent delirium are practical and low technology, but they require coordination. A patient needs glasses, hearing aids, mobilization, hydration, sleep protection, pain control, medication review, and family engagement before delirium appears, not after the first night of agitation. Prevention must therefore begin at admission.

Reframing the Dilemma

The phrase “every elderly admission is a delirium prevention case” is useful if it means that delirium risk should be considered for every older adult. It becomes misleading if it implies that every patient requires the same level of intervention intensity. A robust approach is risk-stratified.

A low-risk 67-year-old admitted overnight for uncomplicated cellulitis may need basic orientation, sleep preservation, medication review, and early mobility. An 87-year-old with dementia, infection, acute kidney injury, hearing impairment, and a Foley catheter needs a high-intensity prevention plan. The hospitalist’s task is to identify the second patient early and prevent modifiable precipitants from accumulating.

Risk Assessment at Admission

Risk assessment should be simple enough to fit real workflows. Core risk factors include age 65 years or older, known or suspected cognitive impairment, dementia, severe illness, hip fracture, sensory impairment, functional impairment, dehydration, malnutrition, polypharmacy, and prior delirium. A medication history is essential because deliriogenic drugs are common and often continued by default during admission.

A useful admission assessment asks four questions. Does the patient have baseline cognitive impairment or dementia? Is the current illness severe or rapidly changing? Are there sensory, mobility, hydration, pain, or sleep vulnerabilities? Are medications likely to worsen cognition or arousal? If the answer to any question is yes, delirium prevention should begin immediately.

Screening and Early Detection

Screening should not be performed unless the hospital has a response plan in place. A positive screen should trigger reassessment for infection, hypoxia, metabolic disturbance, pain, urinary retention, constipation, medication toxicity, withdrawal, dehydration, sleep deprivation, and environmental contributors.

The Confusion Assessment Method is widely validated and useful when clinicians are trained to apply it properly. CAM-ICU is designed for ICU patients, including mechanically ventilated patients. The 4AT is brief and practical for many wards and emergency settings. The best tool is the one the team will use consistently and act on.

Daily observation is as important as formal screening. Nurses, therapists, pharmacists, family members, and bedside sitters may detect early changes before the physician does. New withdrawal, delayed responses, reduced mobility, poor attention, altered sleep-wake cycle, visual hallucinations, agitation, or sudden refusal of care should be treated as possible delirium until proven otherwise.

Prevention: What Actually Belongs in the Bundle

Delirium prevention is not one intervention. It is the cumulative effect of reducing several modifiable stressors.

Orientation should be active. The patient needs visible clocks, calendars, lighting cues, and repeated explanation of location, date, and reason for hospitalization. Family presence can be clinically useful when it improves orientation and reduces fear.

Mobility should begin early unless contraindicated. Bed rest is not benign in older adults. Even sitting upright, transferring to a chair, or performing active range-of-motion exercises may preserve function and reduce the risk of delirium when ambulation is not possible.

Sleep should be protected. Overnight vital signs, laboratory draws, medication administration, alarms, hallway noise, and room changes should be minimized when clinically safe. Sedative-hypnotics are not a substitute for sleep hygiene.

Hydration, nutrition, constipation, and urinary retention should be addressed daily. Dehydration, poor intake, constipation, and bladder distention are common, treatable precipitants of delirium. Indwelling urinary catheters should be avoided or removed as soon as possible unless clearly indicated.

Pain should be treated, but opioid exposure should be deliberate. Untreated pain can precipitate delirium, while excessive opioid dosing can worsen delirium, sedation, constipation, respiratory depression, and falls. Multimodal analgesia should be used when appropriate.

Vision and hearing support should not be optional. Glasses, hearing aids, dentures, and communication aids should be available and used. Sensory deprivation is a modifiable delirium risk.

Medication review should occur at admission and again when mental status changes. The pharmacist’s role is central because delirium risk often rises after medication additions, renal function changes, or drug interactions.

Medication Safety for Hospitalists and Pharmacists

Medication review is one of the highest-yield steps for preventing delirium. Benzodiazepines should generally be avoided in older adults except for specific indications such as alcohol withdrawal, benzodiazepine withdrawal, seizures, selected procedures, severe anxiety when alternatives are inappropriate, or palliative indications. Anticholinergic drugs, sedative-hypnotics, skeletal muscle relaxants, many antihistamines, dopamine blockers, and high-dose opioids deserve scrutiny.

Antipsychotics do not treat the underlying cause of delirium. They may reduce dangerous agitation or severe distress in selected cases, but they should not be used as routine delirium therapy and should not be used for hypoactive delirium. If an antipsychotic is required, the clinician should document the target symptom, use the lowest effective dose, reassess daily, and discontinue as soon as possible.

Haloperidol requires attention to QT prolongation, torsades de pointes, sudden death, extrapyramidal symptoms, neuroleptic malignant syndrome, Parkinson’s disease, Lewy body dementia, electrolyte abnormalities, interacting QT-prolonging drugs, and dementia-related mortality warnings. Atypical antipsychotics share important risks, including sedation, orthostasis, falls, metabolic adverse effects, QT prolongation, stroke risk in dementia, and increased mortality warnings in elderly patients with dementia-related psychosis.

Melatonin and ramelteon should not be presented as routine general-ward delirium prevention. Evidence varies by setting, population, and study design. In ICU practice, recent guidance is more favorable toward melatonin than older guidance, but certainty remains limited. Local protocols should distinguish ICU sedation and sleep-disruption management from general medical ward care.

Table 1. Practical Delirium Prevention Bundle for Hospitalists

Domain	Bedside action	Pharmacy or ordering implication
Mentation	Establish baseline cognition, use family input, and reorient regularly	Avoid deliriogenic drugs; reassess after mental status change
Mobility	Out of bed for meals, ambulate, or use range-of-motion exercises when safe	Avoid unnecessary sedatives, restraints, and immobilizing devices
Sleep	Reduce nighttime disruptions, cluster care, and manage noise and light	Avoid routine hypnotics; time diuretics, steroids, and lab draws thoughtfully
Sensory support	Ensure glasses, hearing aids, dentures, and communication aids	Avoid medications that worsen vision, hearing, or alertness when possible
Hydration and nutrition	Monitor intake, address dehydration, dentures, and swallowing concerns	Adjust renal clearance of drugs; avoid excessive diuresis when possible
Pain	Treat pain and look for nonverbal signs	Use multimodal analgesia; minimize excess opioid burden
Elimination	Prevent constipation and urinary retention; remove Foley catheters early	Review anticholinergics, opioids, bladder agents, and bowel regimen
Infection and hypoxia	Treat infection, assess oxygenation, and avoid unnecessary catheters	Review antimicrobial CNS effects, renal dosing, and interactions
Family engagement	Encourage familiar voices, reorientation, meals, and mobility support	Communicate medication changes and delirium risk clearly

Table 2. Medication Classes That Commonly Worsen Delirium Risk

Drug or class	Why it matters	Practical approach
Benzodiazepines	Sedation, cognitive impairment, falls, paradoxical agitation, and delirium risk	Avoid except for withdrawal, selected procedures, seizures, or palliative indications
Nonbenzodiazepine hypnotics	Sedation, falls, confusion, limited durable sleep benefit	Avoid routine use; prioritize sleep hygiene
Anticholinergics	Impaired attention, urinary retention, constipation, blurred vision, tachycardia	Reduce anticholinergic burden; use alternatives when possible
First-generation antihistamines	Strong anticholinergic and sedating effects	Avoid for sleep, itching, or nausea when safer options exist
Opioids	Sedation, constipation, urinary retention, and respiratory depression	Treat pain, but use the lowest effective dose and bowel regimen
Antipsychotics	Do not reverse delirium; risk of QT prolongation, EPS, falls, stroke, and mortality warning in dementia	Reserve for dangerous agitation or severe distress after non-drug measures fail
Corticosteroids	Insomnia, mood change, psychosis, hyperglycemia	Use the lowest effective dose and avoid unnecessary nighttime dosing
Dopamine agonists and antiparkinsonian drugs	Hallucinations, confusion, orthostasis	Avoid abrupt withdrawal; adjust with neurology input
Polypharmacy after admission	Risk rises when several CNS-active or anticholinergic drugs are added	Require medication reconciliation and pharmacist review

Table 3. Delirium Screening Tools in Hospital Practice

Tool	Best setting	Practical note
CAM	General hospital use when staff are trained	Strong validation, but requires training and attention assessment
CAM-ICU	ICU, including ventilated patients	Useful when speech is limited
4AT	Emergency department, medical wards, surgical wards, and older adults	Brief, practical, and supported by diagnostic meta-analysis
bCAM	Emergency department or rapid bedside assessment	Useful when the workflow requires a very brief screen
RASS	Arousal assessment, especially in the ICU	Not a delirium diagnostic tool by itself

ICU and Surgical Considerations

ICU delirium has additional drivers, including mechanical ventilation, sedatives, immobility, sleep disruption, pain, sepsis, hypoxemia, and organ failure. ICU protocols should incorporate pain control, light sedation, spontaneous awakening and breathing trials when appropriate, delirium monitoring, early mobility, and family engagement. Recent ICU guidance supports enhanced mobilization and selected sedation strategies when delirium reduction is a priority.

Surgical patients require preoperative risk recognition and postoperative prevention. Pain control, early mobilization, avoidance of unnecessary catheters, sleep protection, and medication review remain central. Claims that one anesthesia type universally prevents delirium should be avoided because effects vary by procedure, patient risk, and perioperative context.

Implementation in Real Hospital Medicine

A delirium program fails when it depends on heroic individual effort. It succeeds when prevention is embedded into routine workflows. Admission order sets should include mobility, sensory aids, sleep hygiene, bowel regimen when appropriate, Foley removal prompts, and pharmacist medication review for high-risk patients.

The electronic health record can help, but it should not become another source of alert overload. A useful EHR intervention identifies risk, suggests concrete actions, and assigns responsibility. “High delirium risk” is not enough. The system should clarify who will mobilize the patient, who will review medications, who will contact family, and when screening will occur.

Hospitals with limited staffing can still begin with a focused bundle: identify high-risk patients, remove unnecessary Foley catheters, avoid benzodiazepines and anticholinergics, mobilize early, preserve sleep, ensure glasses and hearing aids are in place, and involve family. These interventions are not glamorous, but they are clinically meaningful.

What to Do When Delirium Occurs

When delirium appears, the first step is not sedation. The first step is to look for causes. Clinicians should assess oxygenation, infection, sepsis, pain, urinary retention, constipation, dehydration, hypoglycemia, electrolyte disorders, renal or hepatic dysfunction, drug toxicity, withdrawal, stroke, seizures, sleep deprivation, and environmental triggers.

Management should match the delirium phenotype and risk. Hypoactive delirium requires recognition, mobilization, hydration, medication review, and treatment of causes. Hyperactive delirium may require de-escalation, family presence, environmental control, and one-to-one observation. Medication is reserved for severe distress or unsafe behavior when nonpharmacologic measures are insufficient.

Restraints should be avoided whenever possible because they can worsen agitation, immobility, injury risk, and delirium. If restraints are used because immediate safety cannot otherwise be maintained, they should be time-limited, reassessed frequently, and paired with active treatment of the underlying cause.

Metrics That Matter

Hospitals should measure both process and outcomes. Useful process measures include admission risk assessment, daily delirium screening in high-risk patients, early mobility documentation, Foley catheter days, anticholinergic burden, benzodiazepine exposure, and pharmacist review completion. Outcome measures include incident delirium, delirium days, falls, restraint use, antipsychotic exposure, length of stay, discharge disposition, readmissions, and family-reported communication quality.

Measurement should not punish clinicians for detecting delirium. Higher detection rates may initially reflect better screening rather than worse care. Quality programs should distinguish incident delirium, prevalent delirium, and documentation improvement.

References